Obstacles to the use of computers in health care have not been computers but rather understanding the complex health care systems in which such machines exist. This application proposes a two-year investigation of technical work in health care. The broad, long-term objectives of this research will be to improve patient safety through three initiatives: (1) develop the behavioral basis for individual and team performance in cognitive artifact development and use, (2) account for implications that behavior has for medical informatics, and (3) develop the problem representation that can be embodied in information applications or devices. Information technology has much in common with analog artifacts as they are often modeled as iconic representations of physical tools. Understanding the origin and nature of the analog artifacts relies on deep and sustained observation of the groups and activities they support in order to derive their meaning. Few information technology projects invest the necessary attention to this understanding. The result is systems that seek to leverage technological capability while being blind to the sophisticated, refined interactions that practitioners have developed to perform complex work. This research will analyze technical work processes involving cognitive artifacts in use at a major urban teaching hospital. This approach is expected to yield insights into the complex dynamics of technical work and the supporting web of distributed cognition. The method has promise for use in other settings, particularly the ICU, ward and emergency room. Cognitive engineering methods will be used to map the distributed cognition processes related to multiple cognitive artifacts. The results of this research will provide a basis to assess other proposals for the use of technology in health care information applications.